# Are LLMs Socially Adaptive? Contrasting Belief Evolution in Large Language Models and Humans

arXiv: https://arxiv.org/abs/2410.10398 · HTML: https://mlsanity.com/paper/2410.10398/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 53 references that are in the map.
- First posted on arXiv 14 Oct 2024.

## Builds on (most cited first, 12 of 12)

- [The Rise and Potential of Large Language Model Based Agents: A Survey](https://mlsanity.com/paper/2309.07864/) (2023)
- [Artificial Intelligence, Values and Alignment](https://mlsanity.com/paper/2001.09768/) (2020)
- [Using cognitive psychology to understand GPT-3](https://mlsanity.com/paper/2206.14576/) (2022)
- [Reward learning from human preferences and demonstrations in Atari](https://mlsanity.com/paper/1811.06521/) (2018)
- [Exploring Large Language Models for Communication Games: An Empirical Study on Werewolf](https://mlsanity.com/paper/2309.04658/) (2023)
- [Agent Hospital: A Simulacrum of Hospital with Evolvable Medical Agents](https://mlsanity.com/paper/2405.02957/) (2024)
- [AgentGym: Evolving Large Language Model-based Agents across Diverse Environments](https://mlsanity.com/paper/2406.04151/) (2024)
- [Deception Abilities Emerged in Large Language Models](https://mlsanity.com/paper/2307.16513/) (2023)
- [Unveiling the Truth and Facilitating Change: Towards Agent-based Large-scale Social Movement Simulation](https://mlsanity.com/paper/2402.16333/) (2024)
- [CompeteAI: Understanding the Competition Dynamics in Large Language Model-based Agents](https://mlsanity.com/paper/2310.17512/) (2023)
- [Do LLM Agents Exhibit Social Behavior?](https://mlsanity.com/paper/2312.15198/) (2023)
- [When Choosing Plausible Alternatives, Clever Hans can be Clever](https://mlsanity.com/paper/1911.00225/) (2019)

## Built on by (most cited first, 11 of 11)

- [Towards a Design Guideline for RPA Evaluation: A Survey of Large Language Model-Based Role-Playing Agents](https://mlsanity.com/paper/2502.13012/) (2025)
- [RhinoInsight: Improving Deep Research through Control Mechanisms for Model Behavior and Context](https://mlsanity.com/paper/2511.18743/) (2025)
- [Interactional Fairness in LLM Multi-Agent Systems: An Evaluation Framework](https://mlsanity.com/paper/2505.12001/) (2025)
- [Do Large Language Models Think Like the Brain? Sentence-Level Evidences from Layer-Wise Embeddings and fMRI](https://mlsanity.com/paper/2505.22563/) (2025)
- [Assessing the Potential of Generative Agents in Crowdsourced Fact-Checking](https://mlsanity.com/paper/2504.19940/) (2025)
- [PICACO: Pluralistic In-Context Value Alignment of LLMs via Total Correlation Optimization](https://mlsanity.com/paper/2507.16679/) (2025)
- [LLMs are Introvert](https://mlsanity.com/paper/2507.05638/) (2025)
- [Bias in Decision-Making for AI's Ethical Dilemmas: A Comparative Study of ChatGPT and Claude](https://mlsanity.com/paper/2501.10484/) (2025)
- [Are LLMs complicated ethical dilemma analyzers?](https://mlsanity.com/paper/2505.08106/) (2025)
- [EMNLP: Educator-role Moral and Normative Large Language Models Profiling](https://mlsanity.com/paper/2508.15250/) (2025)
- [Mitigating Social Sycophancy via Pluralistic Preference Optimization](https://mlsanity.com/paper/2610.02568/) (2026)

## Abstract (arXiv)

As large language models (LLMs) increasingly engage in complex social interactions, ensuring that their behaviors align with human ethical principles and intentions, known as value alignment, has become a critical scientific challenge. Existing benchmarks often rely on static assessments and fail to capture the longitudinal dynamics of decision-making or the latent cognitive processes driving agent behavior. In this work, we propose FairMindSim, a realistic simulation benchmark rooted in social psychology that evaluates alignment through continuous economic games. To move beyond black-box observations, we introduce the Belief-Reward Alignment Behavior Evolution Model (BREM), a probabilistic framework that formalizes decision-making as a dynamic trade-off between maximizing extrinsic rewards and upholding intrinsic beliefs. We conducted a large-scale comparative study involving 1,017 human participants and ten LLMs, including GPT-5 and Gemini-3-Pro. Our experimental results reveal a capability linked non linear empirical trend in the Third Party Punishment (TPP) game. Mid capability models exhibit rigid and algorithmic aggression that is characterized by over punishment, while frontier models show a convergence of restraint and a shift toward human like leniency as reasoning capabilities scale. Furthermore, using BREM, we decompose agents longitudinal decision dynamics and find that more advanced models better balance conflicting objectives by reducing belief action inconsistency. Our contributions provide a standardized protocol for psychological stress testing and an interpretable mechanism for analyzing the longitudinal evolution of AI alignment in controlled social dilemma settings.